Introduction
Diabetic kidney disease is not just the most prevalent form of chronic kidney disease (CKD) but also it is the most leading cause of end-stage kidney disease (ESKD) worldwide [
1‐
4]. Proteinuria, or macroalbuminuria, has been considered to be the clinical hallmark of diabetic kidney disease and an independent risk factor for ESKD [
5,
6]. Patients with diabetic kidney disease are believed to develop proteinuria prior to renal function loss [
7]. This classical belief, however, has been recently disputed by growing evidence that a substantial proportion of patients either with type 1 diabetes or type 2 diabetes have renal function loss in the absence of proteinuria, known as nonproteinuric diabetic kidney disease [
8‐
13]. This phenotype of diabetic kidney disease suggests that there is a dissociation between renal function and level of albuminuria in patients with diabetes and highlight the need for broader understanding of renal function loss apart from those related to an increase in albuminuria. However, a limited number of studies have investigated nonproteinuric diabetic kidney disease.
In this review, we discuss ever known epidemiology, pathology, renal prognosis, and mortality of nonproteinuric diabetic kidney disease, comparing with those of proteinuric diabetic kidney disease. We also discuss potential mechanisms and perspectives of nonproteinuric diabetic kidney disease.
Conclusion and perspectives
Diabetic kidney disease has been clinically diagnosed based on the traditional belief that patients with diabetic kidney disease present proteinuria followed by renal function loss, and glomerular nodular lesions are observed if these patients underwent kidney biopsy. However, a number of clinical and pathological studies of diabetic kidney disease we have reviewed in this compendium have uncovered the evidence that diabetic kidney disease is clinically and pathologically heterogeneous, suggesting that there may be various phenotypes of diabetic kidney disease. One of these phenotypes is nonproteinuric diabetic kidney disease characterized by renal function loss (eGFR < 60 mL/min/1.73 m2) in the absence of proteinuria (UACR ≤ 300 mg/g creatinine or mg/day). This phenotype of diabetic kidney disease suggests that there is a dissociation between renal function and level of albuminuria in patients with diabetes and highlight the need for broader understanding of renal function loss apart from those related to an increase in albuminuria.
A handful of clinical studies raised potential mechanisms of becoming nonproteinuric diabetic kidney disease. One possibility is an increase in elderly patients with diabetes. Senescence of the kidney occurs with aging, which may cause an eGFR below 60 mL/min/1.73 m
2, even if the impact of diabetes on renal function is little. Additionally, elderly patients with diabetes often have increased underlying conditions such as hypertension, dyslipidemia, obesity, and hyperuricemia, all of which may cause renal function loss via arteriosclerosis, known as nephrosclerosis. Indeed, the primary pathological findings in nonproteinuric diabetic kidney disease are similar findings to hypertensive nephrosclerosis, characterized by glomerular sclerosis, interstitial fibrosis and tubular atrophy, and arteriosclerosis [
19,
24,
27‐
29]. Another possibility is that patients with nonproteinuric diabetic kidney disease is mostly comprised of those who responded well to renin-angiotensin system blockades that results in nonproteinuria via protecting glomerulus. A meta-analysis of 28 cohorts including 693,816 patients (80% with diabetes) and 7461 ESKD events has shown that 30% reduction in albuminuria over 2 years was associated with around 20% risk reduction of ESKD, suggesting that regression of albuminuria may be a favorable prognostic indicator [
30]. However, whether nonproteinuric patients finally develop proteinuria and progress to ESKD, despite of multifactorial therapy, is of great interest. Other possibility is macroangiopathy. A couple of studies report that the prevalence of diabetic retinopathy is lower in those with nonproteinuric diabetic kidney disease than those with proteinuric diabetic kidney disease, suggesting microangiopathy may not be the main pathogenic factor, rather past history of macrovascular disease such as cardiovascular disease may be a potential pathogenic factor in nonproteinuric diabetic kidney disease [
23]. However, this mechanism seems doubtful from our results showing that there were no differences in prevalence of retinopathy and CVD events among nonproteinuric and proteinuric diabetic kidney disease [
24,
31].
Although our study has shown that patients with nonproteinuric diabetic kidney disease carry a lower risk of progression of renal function loss, compared to those with proteinuric diabetic kidney disease, around 20% of those with nonproteinuric diabetic kidney disease experienced progression to advanced CKD or ESKD in 10 years [
24]. Those who progressed to advanced CKD or ESKD had more severe interstitial fibrosis and tubular atrophy, compared to those who did not progress, suggesting that in the absence of proteinuria, tubular damage may play an important role in progression of CKD. However, whether those who progressed to advanced CKD finally develop proteinuria needs to be clarified. Data on albuminuria followed up to ESKD is scarce. A study from the Steno Diabetes Center reported that around 20% of diabetic patients in the absence of albuminuria never developed proteinuria before ESKD, suggesting that developing to proteinuria is not a prerequisite for ESKD [
25]. This finding suggests that underlying pathogenesis may different among glomerulus and interstitium. However, this study included patients with clinical diagnosed diabetic kidney disease in single center and therefore it still remains to be elucidated whether nonproteinuric patients finally develop proteinuria and progress to ESKD in other cohorts. Exploration of biomarkers apart from level of albuminuria may elucidate a mechanism of progression of nonproteinuric diabetic kidney disease. For example, an analysis of a nationwide biopsy-based cohort in Japan with a thorough glomerular investigation showed that diffuse lesions, polar vasculosis and subendothelial space widening predict the prognosis of advanced CKD even in the absence of proteinuria [
21]. Notably, the subendothelial space widening also provides prognostic value on predicting CVD events in patients with nonproteinuria. Other examples are that some studies report that potential mechanisms of progression of advanced CKD include inflammation markers such as TNF and Fas pathways, and tubular damage markers such as KIM-1, all of which are reported to be independent of level of albuminuria [
32,
33].
Recent studies focused on nonproteinuric diabetic kidney disease have elucidated its clinical, pathological features, renal prognosis, and mortality. However, further studies are needed to fully comprehend its mechanism and retard its progression of CKD.
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.